ARCHIVES

In the recent era, mental health problems are increasing globally, leading to greater interest in using AI to develop therapy tools. These tools aim to reach and support more people in need. Even with new technology, current AI chatbots often don’t have the emotional understanding, professional knowledge, or flexibility needed to give good mental health support. This paper suggests a new approach that combines prompt design, machine learning, and psychology - especially Cognitive Behavioral Therapy (CBT) - to build AI systems that are more caring and helpful for mental health support. A close look at current research shows four main problems with AI systems: they use general prompts, often misunderstand emotions, forget past conversations quickly, and don’t have strong safety measures for handling crises. To address these issues, the proposed system includes components that generate supportive responses, understand user emotions, apply basic psychological reasoning, ensure user safety, and recall past conversations to provide better support. In conclusion, these features collectively enhance the AI’s ability to provide understanding and emotionally intelligent support, particularly in mental health settings. By integrating both technical and psychological insights, the system aims to offer a user-centered, reliable, and empathetic approach to mental health care.